Abstract—Particle Swarm Optimization (PSO) algorithm is an intelligent optimization algorithm originating in bird predation behavior, which is widely used in functional optimization, neural network training, pattern classification, system control and related field. Although the Particle Swarm Optimization algorithm’s convergence speed is very fast, it is easy to cause premature convergence and poor performance in multi-peak problems. In order to solve these shortcomings, this paper proposed a Cooperative Particle Swarm Optimization algorithm based on normal Cloud Model with cloud mutation operator, and improved it in three aspects. Firstly, the normal cloud mutation operator was added in the process of particles update, which improved the search performance of Particle Swarm Optimization algorithm on multi-peak problems. Second, this paper used the Huffman tree's construction process to divide the particles into different sub-populations, and this method ensured the diversity of the population. Thirdly, the whole encoding process was saved and transmitted by using the paradigm Huffman coding, which reduced the spatial complexity of the algorithm. At the same time, the inertia weight in the algorithm was optimized by the method of adaptive inertia weight and stochastic inertia weight, which balanced the local and global search ability of the particles and ensured the synergy of the algorithm. In this paper, we used the simulation experiment method to conduct 50 independent experiments on four common functions. By comparing the mean of optimal solution, variance, average convergence time, average convergence generation and other indicators of PSO, CH-PSO, HC-PSO and CH-HC-PSO algorithms, a phenomenon was found that the cooperative Particle Swarm Optimization algorithm with cloud mutation operator based on normal Cloud Model is better than the other three algorithms in performance, which effectively solves the problems of premature convergence and multi-peak optimization performance. This algorithm was applied to solve the high-dimensional problems effectively.
Index Terms—Cloud model, cloud mutation operator, paradigm Huffman coding, particle swarm optimization.
J. Luo and Y. Gao are with the School of Computer Science and Education Software, Guangzhou University, 510006 China (Corresponding author: Y. Gao; e-mail: 2111706006@e.gzhu.edu.com; gaoying_gzhu@outlook.com).
Cite: Jiahui Luo and Ying Gao, "Cooperative Particle Swarm Optimization Algorithm with Cloud Mutation Operator Based on Normal Cloud Model," International Journal of Machine Learning and Computing vol. 9, no. 5, pp. 554-560, 2019.
Copyright © 2019 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).